The interneuronal functional associations were studied in two dogs with Nichrome semi-microelectrodes implanted into the deep layers of the motor and somatosensory regions of the cerebral cortex using the method of cross-correlation analysis. For this purpose the impulse activity of individual neurons was distinguished byThe recording of multineuronal activity (MNA) through chronically implanted semi-microelectrodes may be considered one of the most appropriate techniques for the study of the physiological mechanisms of the interneuronal interactions [6]. An increasing number of neurophysiologists are in firm agreement with the notion [7] that the MNA is the impulse activity of neurons which vary in size and consequently are functionally heterogeneous, and which constitute a spatially separate microgroup. Data regarding the nested, columnar organization of the central nervous system serve as the morphological basis for this hypothesis [1,3,4,20].The recording of MNA in combination with cross-correlation analysis, aptly named "electronic horseradish peroxidase" [22], makes it possible to penetrate the structure of actually existing local cortical networks. This is precisely the path which we chose to study the functional organization of microsegments of the sensorimotor region of the cerebral cortex in dogs [8].The impulse series of three neurons were distinguished for these purposes from the background MNA through the use of amplitude discrimination, and their functional associations were studied by means of cross-interval histograms (CIHG), using three fixed bin width values, 0.5, 2, and 10 msec.In awake dogs, both untrained and trained, the cross-interval interrelationships were manifested in the form of single, unilateral peaks, which in terms of form most likely attest to direct excitatory associations between neighboring neurons [11, 211.The complete absence of broad central peaks, which are an indicator of a common excitatory source [21], was surprising, although under anesthesia they were recorded as the basic type of cross-interval association beneath the very same electrodes.Visible indicators of inhibitory interactions, both direct and from a common source, were also absent; this is only partially explained by the lower sensitivity of cross-correlation analysis for the discrimination of inhibitory associations [16].If the cross-interval associations we recorded with short delays up to 10 msec, and delays of medium duration, several tens of msec, could be explained on the basis of theoretical perspectives and facts [21] as manifestations of the direct effects Institute of Higher